Loading
Specialized documentation, mathematical formulations, and empirical backtesting frameworks for quant development & engineering.
Implement and backtest a mean-reversion strategy using the Ornstein-Uhlenbeck process with OLS estimation, half-life calculation, and vectorized signal generation.
Build a sequence-gap-aware, fully asynchronous Level 2 order book consumer for Alpaca Markets. Covers microprice, order book imbalance, reconnection handling, and a realistic latency budget.
Compute defensible 95% confidence intervals for the annualised Sharpe ratio using a stationary block bootstrap. Covers skewness, kurtosis, block-size selection, and the deflated Sharpe ratio correction for multiple testing.
Build a volume-weighted Order Book Imbalance (OBI) signal from L2 market data with adaptive z-score thresholds, decay-weighted multi-level aggregation, microprice anchoring, and a realistic event-driven backtest framework with execution lag simulation.
Design a production-grade event-driven algorithmic trading engine with asyncio event loop, ZeroMQ process separation, heartbeat watchdog, and type-safe event hierarchy. Covers startup ordering, queue backpressure, and clock-sync edge cases.
Build a synchronous pre-trade risk gate in Go with position caps, order rate limiting, notional limits, drawdown circuit breakers, and square-root market impact estimation. Covers clock drift, flapping protection, and race-condition-safe validation.
Detect and eliminate lookahead bias in quantitative backtests with automated linting, purged walk-forward cross-validation, and realistic execution assumptions including slippage, fill probability, and timestamp jitter.
End-to-end latency benchmarks for the tick-to-order hot path across C++, Go (tuned GOGC), Rust, Java (ZGC), and Python (uvloop). Includes methodology, allocation profiles, and a decision framework for when to optimise.
A six-step production deployment path for algorithmic trading strategies: vectorised research, event-driven backtest, realistic execution, paper trading, kill switch deployment, and slow scaling. Includes monitoring, alerting, and edge case handling.
Connect to Interactive Brokers from Python for algorithmic trading. Covers ib_insync connection, Level 2 market data, order placement (LMT, MKT, PEG MID), account tracking, and production resilience with automatic reconnection.
Build a production-grade crypto trading bot in Python for Binance. Covers WebSocket order book streaming, REST authentication (HMAC-SHA256), spread mean-reversion strategy, order reconciliation, and rate-limit handling.
Build a pairs trading strategy with cointegration testing (Engle-Granger, ADF), hedge-ratio estimation, z-score signals, half-life exits, and a walk-forward backtest in Python.
Implement walk-forward optimization in Python: anchored vs rolling windows, re-fit cadence, embargo, stitched out-of-sample equity, and the per-window diagnostics that expose overfit strategies.
The CFTC Commitments of Traders report explained: commercials vs non-commercials, net positioning, extremes vs history, weekly change momentum, and the three mistakes retail traders make.
A five-gate anti-overfitting checklist: trial accounting, deflated Sharpe, purged CV, walk-forward, and point-in-time discipline — with Python code for each gate.